Grazing effects are multi-trophic through mediating plant composition: a meta-analysis



Alessandro Filazzola, Batbaatar Amgaa, Charlotte Brown, Issac Heida, Jessica Grenke, Margarete Dettlaff, Tan Bao, & JC Cahill

library(tidyverse)
library(PRISMAstatement)

Purpose

Conduct a meta-analysis of the literature testing the indirect effects of grazing on animal taxa’s through the direct effects on the plant community.

Objectives

Timeline

date task
Nov 9 Establish search terms to be used in the meta-analysis
Nov 12 Compile list of journal artcles and sub-divide for each researcher
Nov 14 Begin reviewing papers and extracting data
Jan 28 Complete data extraction from papers
Feb 11 Complete preliminary analysis and set structure for MS
Feb 25 Settle on analyses to be used and begin writing manuscript
March 11 Complete first draft of MS and pass to co-authors
March 25 Comments passed back on draft
April 2 Complete revisions and submit to journal

Literature Review - 1. Search

Search Terms

A systematic literature search was conducted using Web of Science for all emperical research articles. The review will include all studies globally. The intended purpose of this search is to capture all articles that have documented grazing either along a gradient (e.g. different frequencies or intensity) or presence/absence (e.g. excluded, ungrazed, or retired ranch lands). We condcuted two separate searches to capture studies that tested gradients and studies that compared grazing to ungrazed treatments. Duplicate articles between the searches were removed. We also intentionally excluded terms that resulted in articles not relevant to the purpose of this study including: review, synthesis, policy, social, carbon, and fish. The search terms that used were:

Search A graz* OR livestock AND exclosure* OR exclusion OR exclude* OR ungrazed OR retire* OR fallow*

Search B grazing intensity OR grazing gradient OR stocking rate OR rotation* grazing

Literature Review - 2. Sort

This steps includes a. checking for duplicating, b. reviewing each instance for relevancy, c. consistently identifying and documenting exclusion criteria. Outcomes include a list of publications to be used for synthesis, a library of pdfs, and a PRISMA report to ensure the worflow is transparent and reproducible. Papers were excluded with the following characteristics:

  • Not emperical study (e.g. review, book chapter)
  • Irrelevant categories (e.g. political science, law, sports tourism, art)

Reasons for exclusion

evidence <- read.csv("data//synthesisdata//evidence.csv")
### Identify studies that were excluded
excludes <- evidence %>% group_by(reason.simplified) %>% count(exclude) %>% filter(reason.simplified!="")
ggplot(excludes, aes(x=reason.simplified, y=n)) + geom_bar(stat="identity") + coord_flip()

## frequency of study
year.rate <- evidence %>% group_by(Publication.Year) %>% summarize(n=length(Publication.Year))

ggplot(tail(year.rate,30)) + geom_bar(aes(x=Publication.Year, y=n), stat="identity") + ylab("number of published studies") +xlab("year published") +theme(text = element_text(size=16))

Prisma report

## total number of papers found
nrow(evidence)
## [1] 2989
## number of papers found outside of WoS
other <- read.csv("data/synthesisdata//other.sources.csv")
nrow(other)
## [1] 0
## number of articles excluded
excludes <- evidence %>% filter(exclude=="yes")
nrow(excludes)
## [1] 2679
## relevant papers
review <- evidence %>% filter(exclude!="yes")
nrow(review)
## [1] 310
## papers for meta
datasets <- read.csv("data//binary.simplified.csv")
meta <- length(unique(datasets$uniqueID))
meta
## [1] 216
prisma(found = 2989,
       found_other = 1,
       no_dupes = 2989,
       screened = 2989,
       screen_exclusions = 2675,
       full_text = 315,
       full_text_exclusions = 0,
       qualitative = 315, 
       quantitative = 216,
       width = 800, height = 800)
## Warning in prisma(found = 2989, found_other = 1, no_dupes = 2989, screened
## = 2989, : After screening exclusions, a different number of remaining full-
## text articles is stated.

3. Synthesis

data <- read.csv("data//binary.simplified.csv")